Marker-assisted selection efficiency in multiple connected populations: a simulation study based on the results of a QTL detection experiment in maize

Many studies already investigated marker-assisted selection (MAS) efficiency but mainly in biparental populations. Connected multiparental populations address a broader diversity and confer a gain of power for QTL detection which must be beneficial for MAS. Our objective was to compare multiparental connected designs to biparental populations taken separately for MAS and phenotypic selection. We first detected QTL for flowering time and grain yield in an experimental maize design involving four parental inbred lines crossed to produce six different biparental populations and confirmed the advantage of multiparental connected designs over biparental populations for QTL detection. Based on these results we performed stochastic simulations to evaluate the expected efficiency of four generations of MAS and phenotypic selection. Different parameters were considered: trait heritability, genetic architecture and whether QTL were assumed to be known or have to be detected. Genetic gains were higher in the multiparental design than on average over the biparental populations considered separately, especially when favourable alleles were equally distributed among parental lines. When QTL detection was included in the simulation process, we found that type I error risk considered for declaring QTL as significant should be adapted to the design. Type I error risks leading to the best response were higher for the biparental populations than for the multiparental design. Besides addressing a broader diversity, multiparental designs increase the power of QTL detection, which reinforces their superiority over biparental designs for MAS. Application of MAS to multiparental designs therefore appears promising to accelerate genetic gain in plant breeding programs.